discretionary v. systematic trader
found in previous post link.
<snip>
I found a website just a minute ago and stumbled upon this article about "Discretionary Trading Vs Systematic Trading" written by Chuck Le Beau...he makes many good points and I feel that he has somehow managed to say what I was thinking (and tried to elaborate on in my earlier post) but in a much clearer manner!..so here it is!
Discretionary Trading Vs Systematic Trading
A recent discussion on the Forum inspired me to select this issue as a subject for a Bulletin. Let's start with my definitions of discretionary and systematic.
I picture a discretionary trader as one who visually analyzes data and arrives at trading decisions with some form of discipline. The term discretionary does not necessarily mean that the trader is shooting from the hip or acting on impulse. A typical discretionary trader tries to arrive at a decision after considering great many factors and probably acts in a fairly predictable pattern that might somehow be computerized given enough time and effort. The typical discretionary trader that I have known looks at chart patterns, a few favorite indicators, and perhaps adds some form of fundamental information and overlays all of this with a personal opinion or bias about the general direction of prices.
I picture a systematic trader as one who has a fixed set of rules that can be implemented by a computer or delegated to a trading assistant for implementation and the trader, the computer and the assistant will all know and agree when a trade should be made.
If the discretionary trader is well disciplined (and most of the successful ones are highly disciplined) it doesn't seem to me that there is a great deal of difference in the two approaches. The discretionary trader can perhaps react better than a computer when there is some important news or unusual market occurrence. The extent of this apparent advantage will, of course, vary from trader to trader depending on their intelligence, knowledge, experience and discipline. In most cases I would expect that a good discretionary trader would out perform a good systematic trader. However, in most cases I would expect the average systematic trader to easily outperform the average discretionary trader.
In fact several years ago I saw a comparison of the performance of discretionary CTAs (commodity trading advisors) and systematic CTAs. The comparison of results over many traders and many years of data clearly favored the systematic traders as a group but a very small number of discretionary traders produced results that were not matched by the systematic traders. The discretionary traders had a very wide range of results varying from terrific to terrible while the systematic trader ranged from pretty good to pretty bad. When the results of each group were combined and averaged the systematic traders had noticeably higher returns than the discretionary traders did. However before we reach any conclusions of our own, lets examine some of the pros and cons of each approach.
The discretionary trader has a great advantage in terms of flexibility and adaptability. The human mind is still capable of analyzing and interpreting vast amounts of data. However the result of that analysis will vary widely depending on the experience and market knowledge of the trader. Emotions and various biases can easily come into play. Fear, hope and greed can distort objectivity. As a group, discretionary traders tend to take profits too quickly and they are often reluctant to take losses in a timely manner.
Other personal factors can also come into the picture. Divorce, illness, a death in the family or other stress-creating situations can obviously impair the decision making process. A discretionary trader can also be unduly influenced by a bad or good series of trades. After six losses in a row the discretionary trader may be reluctant to implement the seventh trade. After six profits in a row the trader may let their guard down and fail to implement their usual thorough analysis and risk control. Unlike systematic traders who can test ideas over historical data discretionary traders have a benchmark problem in terms of expectations. How many losses in a row should a discretionary trader expect? How will a discretionary trader know when to quit or change their strategy? What do you do if you are a discretionary trader and things are going wrong? How big a drawdown should a discretionary trader expect or tolerate?
Systematic trading has many advantages but it also has its share of problems. Mechanical trading systems tend to have a difficult time adapting to changes in markets and one of the few certainties in trading is that markets tend to change over time. There are also potential market disruptions that systems may not be prepared to handle. Market crashes, wars, political events, natural disasters and the weather are just a few of the variables that can temporarily create problems for a systematic trader. Most systems are designed to trade well in a specific type of market environment. For example, a trend following system is unlikely to do well in a non-trending market environment. A counter-trend system is unlikely to do well in a strong trending market.
However systems can be tested over historical data and some idea of trading activity, risk, potential profit and a system's ability to function in various market environments can be crudely estimated. Even though historical testing is not very reliable even when done properly it does give a trader a general idea of what to expect. If for example we know that the system is likely to have eight losses in a row we may not become overly concerned if we have five losses in a row. The system's ability to identify the next trading opportunity is not influenced by the results of the last few trades. Our computer's decision-making ability is not influenced by fear, hope and greed, the flu, or a hang-over. Historical trading results can help us to design money management strategies to maximize profit potential and to control risk.
The choice of trading approaches is a very personal one. Either approach can be made to work or fail. There is some skill and market knowledge required in either method and they are not necessarily mutually exclusive. A trader can easily use both methods. Many discretionary traders run systems and treat them as just another source of information. They may or may not act on the signals based on other factors they wish to consider. Some systems traders switch systems on and off based on their opinions or market outlook or apply some discretion on which trades to skip or implement. |
